JP4756134B2 - Decoding method and apparatus - Google Patents

Description

(Technical field)
The present invention relates to a method and apparatus for decoding data transmitted in a wireless digital communication system.

(Background of the Invention)
The UMTS communication system defined in the UMTS (universal mobile telecommunications system) standard includes at least a node B and a mobile telephone terminal certifying user equipment (UE). Both Node B and UE comprise a transmitter and a receiver. Node B transmits a signal on the downlink to the UE receiver via a Node B transmitter. The UE transmits signals on the uplink to the Node B receiver by the UE transmitter. Several UEs can communicate with a single Node B. In this case, these UEs are said to be in the same cell. A cell represents a geographical area where a UE is served by a given Node B.

  In uplink and downlink communication, both UE and Node B map (associate) information to be transmitted and control data to logical channels. In particular, the UMTS standard provides for downlink communication a logical channel represented by an AICH (acquisition indication channel). When the UE tries to establish communication with a new node, i.e. to join a new cell, the UE asks for permission to become part of this cell. A standard procedure for this purpose uses AICH and is referred to as random access.

  During random access, in order to request access to the cell, the UE sends a short information burst (a chunk of information) called the preamble (preamble), which is a Hadamard code of length 16 Including a signature of the form The UE then waits for the elapsed time to receive a response from the target Node B. This elapsed time is called a guard period. If Node B receives the preamble without error, Node B responds with another burst called acquisition indication (AI).

  If the UE does not receive the AI after a predetermined time, the UE transmits a new preamble. If there is no response from Node B, the UE increases its power and retransmits this preamble up to a maximum number of times. If the maximum number of preamble retransmissions has been reached and normal AI reception has not been achieved, the UE signals this random access procedure failure to a higher network layer.

  The AICH carries a Node B message, ie AI, and acknowledges all UEs attempting to join a particular cell with random access. AI is mapped onto a triplet of characters, which means that AI can take values of 1, 0, and -1. AI is set to 1 or −1 to represent a positive or negative acknowledgment from Node B, respectively. If Node B is unable to provide the service requested by the UE, AI is set to zero.

  All AIs are sent by Node B on a single channel AICH, and all UEs receive all these AIs destined for all UEs. This means that each UE has to decode all information in all AEs even if it is not for this particular UE. Therefore, how to perform efficient decoding of information in the AI becomes a problem.

J. G. Proakis, “Digital communications”, Mc Graw Hill, 3rd Edition

  According to the prior art, standard block code decoding techniques well known to those skilled in the art (eg, JG Proakis, “Digital communications”, Mc Graw Hill, 3rd Edition, pages 436-460) Decryption cannot be performed.

(Summary of Invention)
Accordingly, an object of the present invention is to solve the problem of a method for performing efficient decoding of information in AICH.

The present invention solves this problem by providing:
-Received DS-CDMA (direct sequence-), including iterative calculation of hard decision vectors using decision thresholds having values based on the probability of each triplet character element of each symbol in a hard decision vector code division multiple access: a method of decoding a symbol sequence in a signal; and
A symbol sequence in the received DS-CDMA signal comprising means for iterative calculation of a hard decision vector using a decision threshold having a value based on the probability of each triplet character element of each symbol in the hard decision vector; User device that can be decrypted.

  The method according to the present invention includes decoding of a symbol sequence in a received DS-CDMA signal, and using a decision threshold having a value based on the probability of each triplet character element of each symbol in a hard decision vector. Includes iterative calculation of vectors.

In a preferred embodiment, the method includes multiple steps and includes the following steps:
Demodulating the received signal, thereby providing a symbol sequence;
Calculating a matrix product of the symbol sequence and a Hadamard decoding matrix;
-Calculating an estimated value of the decision threshold assuming an equal probability of the triplet character elements of each symbol in the symbol sequence (equal probability);
Calculating a hard decision vector using the calculated decision threshold;
-Calculating an estimate of the probability of each triplet character element of each symbol in the hard decision vector;
-Calculating a decision threshold using an estimate of the probability of each triplet character element of each symbol in the hard decision vector;
Calculating the hard decision vector until the decision threshold converges or the number of iterations reaches a predetermined maximum number of iterations, estimating the probability of each triplet character element of each symbol in the hard decision vector Calculating a value, and repeating the steps of calculating a decision threshold using an estimate of the probability of each triplet character element of each symbol in the hard decision vector.

  Compared with the prior art, the present invention is very effective. That is, under ideal conditions, ie, in the presence of an adaptive white Gaussian noise (AWGN) channel with normalized signal amplitude (or a priori known signal amplitude), the AICH Standard block decoding may be employed for decoding, but is not as efficient as the present invention, as will be described below.

  Classical Hadamard maximum likelihood (ML) decoding under the AWGN channel is generally performed by multiplying the received code sequence by an Hadamard inverse. The hard decision can be obtained from the resulting decoded sequence.

  For the ML criterion to be applied, it is necessary to calculate a threshold value and define a determination region. In the specific case of AICH where the AI is mapped to the triplet character 1, 0, 1, two thresholds are required. Under an ideal AWGN channel and a known signal amplitude, the threshold can be easily found at 0.5 and -0.5 times this signal amplitude. In practice, these thresholds depend on the actual propagation conditions with the RF-baseband front end of the receiver and the power level of the AICH physical channel that varies with time.

  Although it is feasible to estimate all these parameters at the receiver level, it is computationally dense and not cost effective in terms of implementation. Furthermore, even if an appropriate estimator is implemented in the receiver, a threshold adjustment mechanism should also be implemented, creating additional complexity. Ultimately, the estimation error has a significant impact on the accuracy of the threshold calculation, leading to potentially poor decoder performance.

  Thus, an advantage of the present invention is to provide efficient decoding of AICH by optimally using all available information. A further advantage is that the AICH decoder according to the invention can be implemented in any kind of UMTS receiver and does not require a priori knowledge of the power of the AICH channel, i.e. "blind aspect". Yes, and can be implemented to satisfy practical complexity constraints in real-time systems.

(Preferred embodiment)
Preferred embodiments of the present invention will be described below with reference to the drawings.

  A typical UMTS system 100 is shown schematically in FIG. User equipment (UE) 101 communicates with Node B 103 via uplink 105 and downlink 107. Node B 103 is connected to core network 111 via wireless network controller 109. The core network 111 includes other networks 113 such as other core networks, a mobile communication network (PLMN), a fixed line telephone network (PSTN), a public switched telephone network (PSTN), Alternatively, any kind of data communication network that provides services to any number of devices and subscribers is connected. As will be apparent to those skilled in the art, the system includes almost any number of user terminals represented by terminal 115 and any number of Node Bs and RNCs represented by Node B 117 and RNC (radio network controller) 119, respectively. Can have.

  FIG. 2 shows a user device 200 according to the invention in schematic but somewhat more detail, which user device is suitable for use in the system 100 shown in FIG. 1 and performs the method according to the invention described below. Can do. The user apparatus 200 in FIG. 2 may be any of the UEs 101 and 115 illustrated in FIG.

  The UE is controlled by a main controller (main controller) 209, and the main controller 209 controls operations of the RF transceiver 203, an internal receiver (or channel equalizer (equalizer)) 205, and a channel decoder 207. A memory 211 and an input / output unit 213 are also connected to the controller 209. Connected to the input / output unit 213 are a speaker 215, a microphone 217, a keyboard 219, and a display 221 to provide the user with a convenient means to use the UE when communicating with other users in the system 100.

  Node B 103 transmits a signal to a receiver in the UE on downlink 107 by using the transmitter. The UE transmits a signal to the Node B receiver in the uplink 105 by the UE transmitter.

  In uplink and downlink communication, both UE and Node B 103 map (correlate) data to be transmitted to logical channels. In particular, the UMTS standard provides a logical channel represented on the downlink 107 by an acquisition indicator channel (AICH).

  During downlink communication, the Node B 103 allocates transmission power to the logical channel transmitted to the UE in the Node B cell. The allocated transmit power will be optimized to satisfy some constraints. In practice, the received signal strength is sufficient to support the requested type of service while minimizing the power allocated to a particular UE to minimize interference to other UEs in the same cell. It should be high enough to ensure that it is present (ie to guarantee the link quality necessary to support the service requested by a particular UE). Furthermore, power allocation must also dynamically adapt to changes in propagation conditions, cell load (number of UEs in the same cell, and associated quality requirements), and user service demands. All of these factors affect link quality.

  The UE in the downlink and the inner receiver of the Node B in the uplink each process the received signal to compensate for unwanted effects of the propagation channel and reduce interference to optimize the performance of the outer receiver. . The performance of the outer receiver is strongly dependent on the signal-to-interference ratio (SIR) associated with the signal to be decoded supplied to the output of the inner receiver. Thus, first of all, the goal of the conventional inner receiver is to minimize the SIR at its output to maximize the performance of the outer receiver.

(Data transmission and reception in DS-CDMA)
In order to transmit information for a specific user, the source information data is properly mapped to binary characters, ie bits, by the source encoding technique. And depending on the logical channel under consideration, these bits are protected by using appropriate channel codes to protect the information bits from anything that causes degradation during transmission (noise, jamming). Channel encoding can be performed. In the special case of AICH, there is no channel coding. The channel-coded bits are then a specific modulation scheme (eg, QPSK (quadratic phase shift keying) modulation, well known to those skilled in the art, where 2 bits are mapped to one symbol) Are modulated by mapping these bits into symbols. And in a DS-CDMA system such as UMTS, these symbols are spread over a wider band by filtering with a specific spreading sequence. The spread symbols are usually called chips. Note that the duration of the chip period is smaller than the duration of the symbol period by a factor approximately equal to the bandwidth expansion rate in the spreading operation. This magnification is equal to the number of chips per symbol period and is called the spreading factor.

  Symbol sequences carrying information bits from different sources are spread by different spreading sequences so that multiple users share the same band and the receiver can select the desired data symbol sequence based on the knowledge of the user specific spreading sequences. Can be distinguished and restored. The obtained chip is filtered by a pulse shaping filter (root cosine filter in the case of UMTS) and digital-analog (D / A) converted. The resulting analog signal is then modulated to radio frequency and transmitted on the downlink by the Node B antenna, or transmitted on the uplink by the UE antenna. The radio frequency signal received by the UE antenna in the downlink or the radio frequency signal received by the Node B antenna in the uplink is demodulated and returned to the baseband (or intermediate frequency IF), analog-digital (A / D) is converted to generate a digital baseband signal.

  The UE or Node B receiver processes this burst band signal to recover useful information intended by the concerned user. For this purpose, the receiver estimates the cascade of transmitter chains (cascading) associated with the transmission of useful data, the actual radio propagation channel, and the receiver chain up to A / D conversion. There is a need. In general, this cascade is simply referred to as a channel, and channel estimation generally represents a set of procedures to be performed to estimate a channel.

  Based on the channel estimate, the internal receiver or channel equalizer is set to maximize the contribution of the desired signal and reduce the contribution of the disturbance. In other words, the inner receiver tries to maximize the output SIR. The output of the inner receiver is sent to the outer receiver or channel decoder, which recovers the useful bit sequence by actually undoing the channel encoding operation performed at the transmitter. The higher the SIR at the output of the inner receiver, the better the performance of the outer receiver in terms of signal decoding error probability.

  Below follows a detailed mathematical description of data transmission and reception in DS-CDMA, followed by the mathematical description of AICH decoding, with step-by-step description of the method for realizing AICH decoding, with reference to FIG. End.

(Signal and channel model)
A general DS-CDMA signal and channel model is provided.

は、送信機におけるＤ／Ａ変換後の送信信号を表わし、Ｔ_cはチップ期間を表わし

であり、ここにＭは拡散倍率を表わし、

は床関数の演算子を表わす。これに加えて、シンボル期間をＴ_s＝ＭＴ_cで表わす。さらに、ｓ(ｎ)はｎ番目の変調シンボル（例えばＱＰＳＫ）を表わし、ｄ(ｋ)は拡散シーケンスのｋ番目のチップを表わし、そしてΨ(ｔ)は、システム帯域を制限するパルス整形フィルタを表わす。 (General received signal model)
Consider the general case of transmission of a signal s (t) through a multipath channel with an impulse response h (t, τ). The continuous time complex baseband received signal before A / D conversion is modeled as:
y (t) = x (t) + ν (t) (1)
Here, x (t) represents a portion of the received signal composed of useful data, and ν (t) represents a noise pulse interference term. The signal x (t) is given by:
x (t) = ∫h (t, t−τ) z (τ) dτ (2)
Where h (t, t-τ) represents the time-varying (change with time) channel impulse response,

Represents a transmission signal after D / A conversion in the transmitter, and T _c represents a chip period.

Where M represents the diffusion magnification,

Represents the operator of the floor function. In addition to this, the symbol period is represented by T _s = MT _c . Furthermore, s (n) represents the nth modulation symbol (eg, QPSK), d (k) represents the kth chip of the spreading sequence, and Ψ (t) represents a pulse shaping filter that limits the system bandwidth. Represent.

  Note that this model can be used by systems with several diffusion layers, such as DS-CDMA systems defined by the UMTS, IMT-2000, and IS-95 standards.

ここに、Ｐはマルチパス成分の数を表わし、ｃ_p(ｔ)及びτ_pは、ｐ番目のマルチパス成分に関連する時共変複素係数及び伝播遅延を表わす。マルチパス複素係数ｃ_p(ｔ)は、ＵＥ及びその周辺の散乱物体（例えば建物、丘、しかし他の移動物体も）のノードＢに対する速度に依存する。(１)式で表わされる信号ｙ(ｔ)はＡ／Ｄ変換されて、離散時間信号ｙ[ｉ]＝ｙ(iＴ_c／Ｑ)が生じ、ここにＱはチップレートに対するオーバーサンプリング倍率を表わす。信号ｙ[ｉ]は内受信機によって処理することができ、内受信機は線形の時共変フィルタｆ[ｉ，ｎ]としてモデル化することができ、出力の離散時間信号を次式のシンボルレートで発生する：
ｒ[ｎ]＝α[ｎ]ｓ[ｎ]＋ν[ｎ] (５)
ここで、一般性を失うことなしに、|ｓ[ｎ]|＝１、及び次式を仮定する：

ここに、ｈ[ｉ，ｎ]＝ｈ(iＴ_c／Ｑ，nＴ_s)は、チャンネル及び内受信機フィルタのカスケードの応答を表わし、ν[ｎ]は、内受信機の出力におけるノイズパルス妨害項を表わし、このノイズパルス妨害項は簡単のため、ゼロ平均の複素円対称白色ガウス付加ノイズ（ＡＷＧＮ）としてモデル化する。 Channel h (t, τ) follows a broad stationary steady uncorrelated diffusion model with Rayleigh fading (damping) and multipath response as described in JG Proakis, “Digital communications”, Mc Graw Hill, 3rd Edition, 1995. Assuming that:

Here, P represents the number of multipath components, and c _p (t) and τ _p represent time covariant complex coefficients and propagation delays associated with the p th multipath component. The multipath complex coefficient c _p (t) depends on the velocity of the UE and its surrounding scattered objects (eg buildings, hills, but also other moving objects) relative to Node B. The signal y (t) represented by the equation (1) is A / D converted to produce a discrete time signal y [i] = y (iT _c / Q), where Q represents the oversampling factor with respect to the chip rate. . The signal y [i] can be processed by the inner receiver, which can be modeled as a linear time-covariant filter f [i, n], and the output discrete time signal can be represented by the symbol Occurs at a rate:
r [n] = α [n] s [n] + ν [n] (5)
Here, without loss of generality, assume | s [n] | = 1 and the following:

Where h [i, n] = h (iT _c / Q, nT _s ) represents the response of the cascade of the channel and inner receiver filter, and ν [n] is the noise pulse disturbance at the output of the inner receiver. This noise pulse disturbance term is modeled as a zero-mean complex circular symmetric white Gaussian additive noise (AWGN) for simplicity.

In the following, the superscript (superscript) (•) ^T and (•) ^H represent transposition and Hermitian transposition, respectively.

そして次式のようにブロック符号化され：
ｓ＝Ｈｂ’
ここにＨは１６×１６のアダマール符号化行列を表わす。結果的な符号化シンボルはベクトルから成り、そして前記モデルに従って拡散され、スクランブルされ、そしてパルス整形される。そして、結果的な信号はディジタル−アナログ変換され、無線周波数に変調されて送信される。 (Blind AICH decoder)
(AICH transmission)
Here, AICH decoding will be described. b = [b ₁ ... b ₁₆ ] ^T represents an acquisition indicator column vector (16 × 1) whose elements are b _i ε {0, + 1, −1} for i = 1,. Shall be. The symbols are assumed in exactly the same way.
The symbol consisting of the vector b = [b ₁ ... b ₁₆ ] ^T is first mapped as:

And is block coded as:
s = Hb ′
Here, H represents a 16 × 16 Hadamard coding matrix. The resulting encoded symbol consists of a vector and is spread, scrambled and pulse shaped according to the model. The resulting signal is then digital-analog converted, modulated to a radio frequency and transmitted.

(AWGN channel)
Assume a white Gaussian additive noise channel with flat and slow fading. In the case of fading, the coherence time is assumed to be greater than 2 slots. The transmitted signal s is spread over 4096 chips (less than 2 slots) and the channel is consequently considered to be flat on the observation window.

に従うν[ｎ]を伴う。αはいくつかの寄与分、即ちＡＩＣＨ物理チャンネルのパワーレベル、伝播条件、受信したＲＦからベースバンドへの変換ゲイン、及び内受信機のインパルス応答をモデル化する実ゲイン（利得）率である。 At the output of the internal receiver, the signal is:
r [n] = αs [n] + ν [n]
Complex circular symmetric Gaussian distribution

Accompanied by ν [n]. α is a number of contributions: AICH physical channel power level, propagation conditions, received RF to baseband conversion gain, and real gain (gain) factor that models the impulse response of the inner receiver.

(demodulation)
The demodulated signal r ′ (16 × 1) is given by:

が得られる。ｇは、復号化行列の軟判定（ソフト・デシジョン）値出力を表わすものとし、即ち次式の通りであり：

ここに

は、アクイジション・インジケータ・ベクトルｂの推定値であり、結果的に、次式のようになる：

(Blind AICH decoder)
Multiply r ′ by the decoding matrix, ie H ⁻¹ . Due to the properties of the Hadamard matrix,

Is obtained. Let g denote the soft decision value output of the decoding matrix, ie:

here

Is an estimate of the acquisition indicator vector b, resulting in:

  The “blind” aspect comes from the lack of a priori knowledge of α. In general, this algorithm can be applied without knowledge of the power level of the AICH channel and without an estimate of the total receiver gain (radio module gain + internal receiver gain). Conversely, if this algorithm is implemented in the UE, the Node B does not need to send AICH power information (avoid some overhead).

に従う。
ブラインドＡＩＣＨデコーダはこの関係を用いて、

及び

を共に推定する。このことは、システムが「凍結する」（推定値

及び

が安定になる）まで反復的な方法で行うことができる。 Equation 6 leads to an estimated model:
g = α · b + η
Where η is a real Gaussian distribution

Follow.
The blind AICH decoder uses this relationship to

as well as

Are estimated together. This means that the system “freezes” (estimated

as well as

Until it becomes stable).

In the following, the problem is considered symmetric by AWGN with zero mean and a priori probability (1/3) of each element in b.
As a result, the determination method can be as shown in FIG. 4, where the determination threshold is T = (α / 2).
For example, consider the operation in iteration ():
K ₋ : Number of strictly negative elements in g K ₊ : Number of positive (or 0) elements in g n ₀ : Number of occurrences of event “g _i = 0” n−1: Event “g _i = −1 If the number of occurrences of i is, for example, g _i <0 (if at least one negative number is present in g), then:

と称する。ｉが例えばｇ_i≧０であれば（ｇ内に少なくとも１つの正値が存在するものと仮定すれば）、次式となる：

は反復０回における推定値を表わすものとする：
場合１：Ｋ_-及びＫ₊が０とは異なる

場合２：Ｋ₊のみが０とは異なる

場合３：Ｋ_-のみが０とは異なる

Since this estimate is symmetric, the same is true for positive elements in g, and another estimate of α is obtained in 0 iterations,

Called. If i is, for example, g _i ≧ 0 (assuming that there is at least one positive value in g), then:

Let denote the estimate at 0 iterations:
Case 1: K ₋ and K ₊ are different from 0

Case 2: Only K ₊ is different from 0

If 3: K _- different from the only 0

の第１推定値があり、図４に示す判定方式を適用すれば、次式の規則（ルール）に従うｂの第１推定値

が得られる：
ｉ＝０,...,１５について：

here,

If the determination method shown in FIG. 4 is applied, the first estimated value of b in accordance with the following rule (rule):

Will yield:
For i = 0, ..., 15:

が得られ、そしてこれを用いて

が得られる。しかし、主な違いは、

は

の関数となる、ということにある。 Here, an operation in k iterations (k> 0) is considered.
As with 0 iterations, estimate

And using this

Is obtained. But the main difference is

Is

It is to be a function of

ｉが例えばｇ_i＜０であり、

の場合には、次式となる：

ｉが例えばｇ_i＞０であり、

の場合には、次式となる：

そして次式が得られる：

Some variables need to be defined in k iterations:

i is for example g _i <0,

In the case of:

i is for example g _i > 0,

In the case of:

And the following formula is obtained:

が

の関数として得られる。 Finally, using the same decision rule as zero iterations, that is, the rule described by Equation (7),

But

As a function of

が収束するまで（即ち、システムが「凍結」するまで）反復すべきである。一般に３回の反復で十分であることがシミュレーションによって示されている。 In theory,

Should be iterated until they converge (i.e., until the system "freezes"). Simulations have shown that three iterations are generally sufficient.

の推定値の中から、特定のＵＥに関係する唯一のＡＩを選択することから成る。このＡＩの指標（インデックス）はノードＢによって提供される。 And the final step is

And selecting a unique AI related to a specific UE from the estimated values. The index (index) of this AI is provided by the node B.

  Returning to FIG. The algorithm described mathematically above is described by a plurality of steps in the method. This method is preferably implemented by software instructions in the control unit of the user equipment, eg user equipment 100, 200 described above with reference to FIGS.

  This method is suitable for implementation in any UMTS baseband receiver, for example a regular rake receiver (rake receivers are well known to those skilled in the art).

  The method includes a preliminary calculation, which includes providing a model of the propagation channel for AICH, which model includes a finite number P of isolated multipath components that follow a broad stationary model of uncorrelated scattering. It is a linear superposition and this multipath component is characterized by time-covariant complex coefficients and delays.

  A further preliminary calculation is to construct a discrete time version of the received signal corresponding to the AICH logical channel at the output of the inner receiver.

And AICH decoding is as follows:
In the demodulation step 301, the symbol modulation of the received signal at the output of the inner receiver corresponding to AICH is demodulated (reverted).
In the multiplication step 302, the matrix product of the signal obtained in the demodulation step 301 and the Hadamard decoding matrix is calculated.
In the estimation step 303, the judgment threshold value is estimated assuming the equal probability of each AI triplet character element {-1, 0, 1}.
In the calculation step 304, the hard decision (-1, 0 or 1) on the output of the multiplication step 302 is calculated using the threshold value calculated in the estimation step 303.
In the estimation step 305, the probability of each element {-1, 0, 1} in the hard decision vector obtained in the calculation step 304 is estimated.
-In the calculation step 306, a new determination threshold value is calculated. Instead of assuming an equal probability of each AI character element as in the estimation step 303, the probability estimated in the estimation step 305 is used.
In decision step 307, if the decision threshold converges to a fixed value or if the maximum number of iterations has been reached, whether to repeat steps 304 to 307 or cancel the iteration process and select step 308 Determine if it should be executed.
-In the selection step 308, from the hard decision vector elements, select a unique AI related to the UE (an indicator of this AI is given by the Node B).

  The present invention is not limited to the above-described embodiments, and changes and modifications can be made without departing from the scope of the present invention defined in the claims.

  In particular, the present invention is not limited to the UMTS applications described above. The present invention can be used within DS-CDMA or other wireless communication system applications where the communication standard is intended for the existence of logical channels containing symbols or signal waveforms that are unknown to the receiver.

  Note that the method according to the invention is not limited to the implementation described above. Given that a single item of hardware or software can perform several functions, there are numerous ways to implement the functionality of the method according to the invention with items of hardware and / or software. The present invention does not exclude that an assembly of items of hardware and / or software performs a function. For example, the algorithm steps can be combined to form a single function with transmit beamforming according to the present invention without changing the method of channel estimation.

  The hardware and / or software items can be realized in several ways, for example by means of individually wired electronic circuits or by appropriately programmed integrated circuits. This integrated circuit becomes part of a receiver that can be included in a computer, in a mobile communication handset (handset), in a base station (base station), or in any other communication system device. The receiver comprises means of a receiver configured to perform the operations necessary to support a particular type of communication service (ie descrambling (de-scramble), de-spreading). This means is a hardware or software item as described above.

  As described above, the means may be hardware or software that performs a function, or an assembly of both, or a single item that performs several functions.

  The integrated circuit comprises a set of instructions. Thus, for example, these sets of instructions contained in a computer programming memory or in a separate receiver memory cause the computer or receiver to perform different steps of the estimation method.

  These sets of instructions can be loaded into the programming memory by reading a data carrier, such as a disk. The service provider may also make these sets of instructions available via a communication network, such as the Internet.

  Apparently, “comprising” and its conjugations do not exclude the presence of steps or elements other than those specified in the claims. Any step or element recited in the claims does not exclude the presence of a plurality of these steps or elements.

1 schematically shows a system including a communication terminal according to the present invention. 1 schematically shows a communication terminal according to the invention. 4 is a flowchart of a method according to the present invention. It is a figure which shows the determination area | region of soft AICH decoding.

Claims (4)

A method for decoding a symbol sequence in the received DS-CDMA signal,
In a method comprising iterative calculation of the hard decision vector with a decision threshold having a value based on the probability of each triplet character element of each of the symbols in a hard decision vector ,
Demodulating the received signal, thereby generating a symbol sequence;
Calculating a matrix product of the symbol sequence and a Hadamard decoding matrix;
Assuming an equal probability of the triplet character element of each symbol in the symbol string, and calculating an estimated value of the decision threshold;
Calculating a hard decision vector using the calculated decision threshold estimate;
Calculating an estimate of the probability of each triplet character element of each of the symbols in the hard decision vector;
Calculating a decision threshold using an estimate of the probability of each triplet character element of each of the symbols in the hard decision vector;
Calculating the hard decision vector; calculating an estimate of the probability of each triplet character element of each of the symbols in the hard decision vector; and each triplet character element of each of the symbols in the hard decision vector. Repeating the step of calculating a decision threshold using an estimate of the probability of until the decision threshold calculation converges or the number of iterations reaches a predetermined maximum number of iterations;
A method for decoding a symbol sequence , comprising : The symbol sequence is a sequence of acquisition indicators (AI) in the acquisition indicator channel (AICH), and further:
The method of claim 1, comprising selecting a related AI from the calculated hard decision vector using a predetermined index. A user apparatus capable of decoding a symbol sequence in a received DS-CDMA signal, wherein a decision threshold having a value based on a probability of each triplet character element of each symbol in a hard decision vector is used. In a user device comprising means for iterative calculation of said hard decision vector,
The means for iterative calculation of the hard decision vector is:
Means for demodulating the received signal, thereby generating a symbol sequence;
Means for calculating a matrix product of the symbol sequence and a Hadamard decoding matrix;
Means for calculating an estimated value of the decision threshold assuming an equal probability of the triplet character element of each symbol in the symbol sequence;
Means for calculating a hard decision vector using the calculated estimated threshold value;
Means for calculating an estimate of the probability of each said triplet character element of each said symbol in said hard decision vector;
Means for calculating a decision threshold using an estimate of the probability of each of the triplet character elements of each of the symbols in the hard decision vector;
Means for calculating the hard decision vector; means for calculating an estimate of the probability of each triplet character element of each symbol in the hard decision vector; and each triplet character element of each symbol in the hard decision vector. Means for calculating a determination threshold value using the estimated value of the probability of repetitive operation until the calculation of the determination threshold value converges or the number of iterations reaches a predetermined maximum number of iterations;
A user device comprising: The symbol sequence is a sequence of acquisition indicators (AI) in the acquisition indicator channel (AICH), and further:
The user apparatus according to claim 3, further comprising means for selecting a related AI from the calculated hard decision vector using a predetermined index.

Images